SKF AWAIV-12, User Manual

Page: 151 / 250

Determining faults
If the insulation is weak or failing on a particular bar or coil of the armature, the test wave
pattern will become unstable and shift left when the section that contains the fault passes
through the “test area.” The test wave pattern will no longer match the reference wave
pattern. This indicates shorted windings within the span.
Usually, as soon as the bad bar is placed under the hot brush/test lead, the wave pattern will
show the shift to the left as noted above. Thus, the bar directly below the hot brush/test lead
is the faulty bar. The illustration below shows an example of a fault discovered after
performing a surge test using the motor’s brushes. On the left, we see the results of a good
surge result, while on the right we see the results for a fault grounded under the brush or
outside of surge test span.
Figure 123.
Comparing good (left) and short results (right).
Usually, as soon as the bad bar is placed under the hot brush or test lead, the wave pattern
will show the shift as noted above. Given the very parallel nature of armatures, the
waveforms will start changing (with respect to the reference) as soon as testing gets close to
the shorted bar. Over the short, the waveform shift will be the greatest. When the ground
brush or probe is over the short, there will still be a shift, but not as great as when the hot
brush/probe is over the short.
Notes and tips for span tests of armatures
• A test fixture may be used instead of using the motor’s brushes to make contact with
the armature.
• Set the span between the fixture’s brushes to the desired number of commutator
bars. Either move the fixture around the commutator during testing or rotate the
armature. Procedures for testing and fault determination are the same.
• Always HiPot the armature to ground first. This gives an upper limit for the
maximum voltage to apply when surge testing.
• The greater the span surge test voltage is, the more adequate the stress between
bars is (ideally, 375 V according to Paschen’s Law). Voltage stress is measured by the
differential or drop between each bar. For example, a ten-bar span with 1,000 V
applied to it will result in a 100 V stress between bars. If the span is lowered to five
bars, 1,000 V applied to the span will result in 200 V between bars.
• Consider, however, that a ten-bar span at 375 V between bars would require a span
test voltage of 3750 V. This potential to ground at the first coil may be too high. A
lower span test voltage is recommended if, for instance, the HiPot test was only to
2,200 V.
SKF Static Motor Analyzer—Baker AWA-IV User Manual 137
Motor testing theory and reference